box details + side profile
closed box, banana for scale
the Markov Chain mathematical system was our main form of inspiration. We saw how for each state within a state space, there was a quantifiable potential in the number that was produced by the probability of transitioning from one state to another. We saw how this could be applied to real life tracking situations where people went from one location to another, and immediately saw surveillance as a practical use. However, it was obviously an ethically grey issue, so we chose to follow the user-tracking theme and find a method of tracking that would provide data that was beneficial to the user, and could not be collected as way to market or take advantage of the user. We wanted to create a project for social good.
What it does
Inspired by the Markov Chain mathematical system, Habit@t combats disruptive habits by gathering real-time data from multi-tracking sensors placed around a person’s home or problem area, which allows the user to have a tangible track record of where/how a habit starts/ends at that location. It is intended to be used as a medical evaluation tool.
One of its main practical applications is to help manage general dysregulated behaviors (eg. excess drinking, binge eating, self-harm). We designed based on the premise of "watch what people really do, and not simply trust what they tell you they do": we want to allow users to view the pathway and consequences of their behaviors without personal bias, in a quantitative way that can be analysed by both medical professionals and civilians to help with diagnoses and treatment. It also allows individual tailoring the placement of sensors, as well as amount used, by doctors to each patient’s unique situation.
How we built it
GCP: Neo4j for graphing database, + Node.js, front end frameworks IBM Watson powers our node-server interaction, IoT integrations with Telus LTE-M IoT starter kit. We 3D printed a casing for the hardware using PLA filament on a Prusa i3 MK3 printer. CAD software used was Autodesk Inventor.
Challenges we ran into
Initially, we had trouble converting Markov chain theory from a set of equations to a set of actionable steps, able to be understood by the everyday civilian. Eventually, we boiled it down to its most basic components so that it was easily executable by our programmers, as well as easy to explain it to someone in a minute in an elevator pitch.
Accomplishments that we're proud of
We created a user-friendly hardware system that could seamlessly blend in to someone's home, as well as be assembled by the user themselves to save on costs. we're proud of the fact that we focused on more than practicality and functionality by keeping creating user personas during our idealization process to get a better idea of the people we are creating our product for.
What we learned
This whole experience was quite a steep learning curve for us. We challenged ourselves to try something new and use the TELUS IoT, as well as attempting to use more than one type of sensor per application.
What's next for Habit@t
Over time, a hypothetical future for Habit@t would be the creation of a smaller, leaner sensor, so it would reduce production costs and blend in more seamlessly into living spaces.
If initial testing for the product turns out well, another hypothetical future is to begin pitching to VC firms, and distribute them across hospitals and clinics. They can also be used as a learning tool for university and high school students, to learn about real-life applications of Markov Chains and inspire the future generation.